Research map / Architectures and efficiency / CNN design
Visualizing Deep Convolutional Neural Networks Using Natural Pre-images
Aravindh Mahendran, Andrea Vedaldi
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use the proposed regularized optimization instead of HOGgle | HOG inversion · ILSVRC 2012 val | reconstruction error (%): 60.1 -> 36.6 | Improved |
| use bilinear gradient orientation assignment as in SIFT (ablation) | HOG inversion · ILSVRC 2012 val | reconstruction error (%): 36.6 -> 11.5 | Improved |
| use stronger regularization (small C) (ablation) | AlexNet inversion, early layers · ILSVRC 2012 val | reconstruction quality | Hurt |
| use stronger regularization (small C) (ablation) | AlexNet inversion, deep layers · ILSVRC 2012 val | reconstruction quality | No clear effect |
| add at least weak regularization (C=100 or C=1) vs none (ablation) | AlexNet pre-images · user study, 13 subjects | interpretability preference | Improved |
| use strong regularization (C=1) on deep layers (ablation) | AlexNet pre-images · user study, 13 subjects | interpretability preference | Improved |
| replace TV (beta=1) with TV^beta (beta=1.5 or 2) (ablation) | pre-image regularizer · AlexNet relu3 inversion | image quality | Mixed |
| add jitter regularizer (ablation) | activation maximization · AlexNet fc8 tree frog neuron | visual structure | Improved |
Closest papers by meaning
- Inverting Visual Representations With Convolutional Networks (CVPR 2016 · similarity 0.90)
- Understanding Image Representations by Measuring Their Equivariance and Equivalence (IJCV 2018 · similarity 0.88)
- A Powerful Generative Model Using Random Weights for the Deep Image Representation (NeurIPS 2016 · similarity 0.88)
- Discriminating image representations with principal distortions (ICLR 2025 · similarity 0.86)
- Towards flexible perception with visual memory (ICML 2025 · similarity 0.86)
- IDEAL: In-DEpth ALignment Makes A Discrete Representation AutoEncoder (NeurIPS 2026 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
